Browsing by Author "Misra, Sushreyo"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Seismic Resilience of Rail-Truck Intermodal Freight Transportation Networks
    (2020-08-14) Misra, Sushreyo; Padgett, Jamie Ellen
    Railway and highway networks constitute the backbone of the US freight transportation network, and the rail-truck intermodal combination constitutes a popular emerging mode of freight transport. Although components of the intermodal network, namely railway bridges, highway bridges, roadways, railway tracks and intermodal terminals have suffered damage in past earthquakes with potentially significant economic consequences, a framework for assessing intermodal network resilience incorporating key component level input models is lacking in the literature. This study introduces a framework for quantifying the time evolving functionality and consequently resilience of rail-truck intermodal freight transportation networks subjected to seismic hazard, incorporating input datasets and models that support the framework. The intermodal network is modeled as an integrated multi-scale network, enabling explicit modeling of network component disruptions on a high-resolution local scale near the site of the disruption event, as well as modeling resulting network throughput on a nationwide scale. In addition to formulating the overarching framework for resilience modeling of intermodal transportation networks, this thesis addresses pressing gaps in modeling the fragility and restoration of constituent components of these systems. Fragility models offer conditional probabilities of physical damage given the intensity of the hazard as well as other structural parameters, offering key input to overall resilience assessment of these networks. A new fragility modeling approach is proposed leveraging elastic nets regularization and logistic regression, and given that they are altogether lacking in the literature, this method is applied to derive new fragility models for typical railway bridge classes subjected to seismic hazards. Restoration models used in the resilience modeling framework, providing estimates of closure decisions and durations given damage states of intermodal network components, are scarce in the literature. Those that exist suffer from the use of limited expert opinion data and lack sufficient insights to relate practical estimates of closure to functionality. To this end, new restoration models are proposed for network components leveraging decision trees and clustered random forests, estimating both decisions and durations of complete closure as well as partial closures (e.g. speed restriction and load restriction). In addition to this, a fault tree model is proposed to model intermodal terminal functionality, enabling integrated assessment of rail and highway networks including explicit modeling of the performance of the nodes of freight transfer. Finally, a restoration scheduling strategy is proposed for optimal allocation of repair crew and corresponding network flows under limited resources, aiming to minimize costs from various stakeholders’ perspectives while ensuring shipment demands are satisfied as far as possible. The framework and input models are tested using a case study analysis on the intermodal network of Memphis, TN subjected to an earthquake originating in the nearby New Madrid Seismic Zone. Overall, this thesis provides a framework for estimating the resilience of intermodal freight networks, while addressing gaps in key input models required to support the framework. The proposed framework and input models will be integrated within Interdependent Networked Community Resilience Environment (INCORE), an open source tool for community resilience modeling currently in development within National Institute of Standard and Technology (NIST) funded Center of Excellence in Community Resilience Planning. As illustrated in the case study application, this framework allows exploration of central questions in infrastructure resilience assessment, such as the spatial distribution of damage and relative impact of various hazard events; the temporal evolution of component and network level performance; the probability distribution of alternative resilience metrics specific to intermodal freight networks; or the impact of different approaches to restoration scheduling and post-event resource deployment. Furthermore, the models posed herein form a basis for probing broader questions in community resilience planning and decision-making, where the resilience of intermodal transportation infrastructure can have major implications on economic or social systems modeling given their role in goods transport, business activity and employment, and recovery of a community.
  • Thumbnail Image
    Item
    The use of machine-learning methods for post-earthquake building usability assessment: A predictive model for seismic-risk impact analyses
    (Elsevier, 2023) Tocchi, Gabriella; Misra, Sushreyo; Padgett, Jamie E.; Polese, Maria; Di Ludovico, Marco
    The assessment of building usability in the aftermath of an earthquake is mostly aimed at post-event emergency management, but it is also valuable for the planning of risk-reduction policies. In the seismic risk assessment field, the development of suitable consequence functions that correlate physical damage to usability and serviceability of structures is crucial to evaluate the expected social and economic losses in a region of interest. Predictive models for usability classification generally are calibrated on empirical data and provide the probability of loss of usability as function of the intensity measure, the building type and the severity of damage attained by the structure. Exploiting the large amount of data available in Italy, a decision tree-based approach is proposed in this study to assess post-earthquake usability of ordinary buildings. Thanks to its high interpretability coupled with reasonable predictive capability _, the selected machine learning algorithm allows investigation of the structural parameters that have a significant impact on building usability, while also accounting for the traditionally neglected uncertainty of subjective decisions. Finally, to show the potential of the proposed usability consequence models, a large-scale risk analysis is carried out to evaluate the spatial distribution of expected building-usability losses over time.